Heat capacities of organic compounds in solution. II

II. Tetraalkylammonium bromides. Edward M. Arnett, and James J. Campion. J. Am. Chem. ... J. A. Burns , R. E. Verrall. Journal of Solution Chemistry 1...
0 downloads 0 Views 571KB Size
7097

then the implications for mechanistic studies are startling, and we will have to completely reexamine our definition of diradical intermediates and our simplistic view of nonconcerted reactions. l9 Acknowledgment. We are indebted t o A. Imamura, W. J. Hehre, and J. Lisle for computational assistance (19) Another area where we may have to face an intermediate simulated by a flat-topped potential surface is in some nonclassical carbonium ions.

and discussions. This work was supported by the National Institutes of Health (GM 13468) and the National Science Foundation (GP 80 13). Acknowledgement is made to the donors of the Petroleum Research Fund, administered by the American Chemical Society, for partial support of this research. The stay of one of us (R. G.) a t Cornell was made possible by a grant from the Deutsche Forschungsgemeinschaft.

Heat Capacities of Organic Compounds in Solution. 11. Some Tetraalkylammonium Bromides’ Edward M. Arnett* and James J. Campion

Contribution from the University of Pittsburgh and Mellon Institute of Carnegie-Mellon University, Pittsburgh, Pennsylvania 15213. Received October 9, 1969 Abstract: Partial molar heat capacities of solution ACpzoare reported for a series of tetraalkylammonium bromides. The process is transfer from the pure solid state to high dilution in water as measured by the integral heat method of Criss and Cobble. Heats of solution are shown to be extremely dependent upon the hydrophobic structure of the ion and upon the temperature. This gives warning against interpreting heats of solution for compounds of this type at a single temperature in simple terms. The corresponding heat capacity of solution for tetran-butylammonium bromide in ethanol is reported and results contrasting to those in water are discussed.

F

or a variety of good reasons, tetraalkylammonium salts occupy a special position in the development of modern solution theory. As a result they have been subjected t o unusually extensive and systematic scrutiny by a battery of physical methods, with particular emphasis on their behavior in aqueous solution.2 This combined effort has been described by Franks3 as “a textbook example of how. . . water-solute interactions should be studied.” The very large partial molal heat capacity of tetra-n-butylammonium bromide in aqueous solution reported by Frank and Wen4 and recently confirmed by Ackermanj is often cited a s powerful evidence that nonpolar groups enhance a degree of temperature-dependent structuredness on water molecules close t o them. In our previous paper6 concerning the determination of heat capacities of solution for a series of alcohols in water,’ we described the application of Criss and Cobbles “integral heat of solution” method. A simple solution calorimeter described by us9 was

used t o measure the temperature coefficient of the partial molal heat of solution (AfTS) for the solute from a pure liquid or solid state to high dilution in water. It, therefore, gives directly the difference in heat capacity (AC,,’) between the heat capacity of the pure solute [as a liquid (CPL) or gas (CPG) or solid (C,’)], and that of its partial molal heat capacity at infinite dilution, This is defined formally as

cp2.

cp2

AC,,’

=

-

Cp, - CpAL’ G , Or

(1)

In this paper, we will describe the result of applying this method t o a series of crystalline tetraalkylammonium bromides in water a t temperatures mostly between 10 and 50”. Our results may be compared with those from a similar study (published since the present paper was first submitted) by Sarma, Mohanty, and Ahluwalia, lo which covers a wider series of salts but which was limited to two temperatures. Their paper also provides a bibliographical entrke t o the field.

*

To whom correspondence should be addressed. (1) This work was supported by a grant from the Office of Saline Water, U. S. Department of the Interior, (2) J. E. Prue, A. J. Read, and G. Romeo in “Hydrogen-Bonded Solvent Sjstems,” A. K . Covington and P. Jones, Ed., Taylor and Francis, London, 1968, p 1 5 5 . (3) F. Franks, see ref 2, p 40. (4) H. S. Frank and W. Y . Wen, Discussions Faraday SOC.,24, 133 ( 1957). ( 5 ) H. Ruterjans, F. Schreiner, U. Sage, and Th. Ackerman, J . Phys. Chem., 73, 986 (1969). (6) E. M . Arnett, W. B. Kover, and J. V. Carter, J . Amer. Chem. SOC., 91, 4028 (1969). ( 7 ) D. M . Alexander and D. J. T. Hill, Ausr. J . Chem., 22, 347 (1969). (8) C. M. Criss and J. W. Cobble, J . Amer. Chem. SOC.,83, 3223 (1961). (9) E. M. Arnett, W. G . Bentrude, J. J . Burke, and P. M. Duggleby, ibid., 87, 1541 (1965).

Experimental Section Preparation and Purification of Chemicals. With the exception of tetraisoamylammonium bromide [(i-Am)rNBr], the tetraalkylamonium bromides and other reactants were of Eastman White Label quality. Tetramethylammonium bromide ( Me4NBr) was

recrystallized once from water. Tetraethylammonium bromide (Et4NBr) was recrystallized once from isopropyl alcohol. Tetran-propylammonium bromide (Pr4NBr), tetra-n-butylammonium bromide (BurNBr), and [(i-Am)4NBr]were recrystallized once from a chloroform-ether mixture (10-15 volumes of ether added to the

chloroform).

(IO) T. S. Sarrna, R. K. Mohanty, and J. C. Ahluwalia, Tram. Faraday SOC.,62, 2333 (1969).

Arnett, Campion 1 Tetraalkylammonium Bromides

7098 [(i-Am)4NBr] was prepared by refluxing equimolar mixtures of 1-bromo-3-methylbutane and triisoamylamine in acetonitrile for 48 hr. The product was crystallized by the addition of ether. The salt was then recrystallized from a chloroform-ether mixture. All the tetraalkylammonium bromides were dried in a vacuum oven at 50-60” for at least 24 hr. Sodium chloride (Baker Analyzed Reagent) was dried at 130” and was used with no further purification. The purities of the tetraalkylammonium bromides were checked by means of a potentiometric titration with standard A g N 0 3 using a silver indicator electrode and a glass electrode as reference.” The potential was monitored with a Beckman Model D R pH meter. Approximately 4 mmol of the bromide salt was dissolved in cu. M H N 0 3 and titrated with 0.100 M AgN03. Potential reddings were taken at 0.50-ml increments near the equivalence point, and the end point was determined by the “analytical” method.I2 Analysis for these salts based on the correct formula weight were in the range of 99.7-100.3Z compared to a test run with potassium bromide (Baker Analyzed Reagent) which gave analyses of 100.9 and 101.3%. Calorimetry. A solution calorimeter was employed which was essentially the same as the one which we described e l ~ e w h e r e , ~ but with the following exceptions. The 2-K resistor of the complementary arm of the Wheatstone bridge was replaced by the thermistor of a twin calorimeter, instead of the usual base-line compensator. A Kepco Model ABC 40-0.5(M) power supply replaced the 6-V batteries. The 200-ml dewar flasks were placed in brass jackets for immersion in a 10-gal aquarium water bath. The Teflon head had two more holes drilled for the accommodation of a total of three solid-injection syringes. Temperatures above ambient were maintained with a no. 20 constant temperature circulator (Bronwill Scientific Division/Will Corp.). Temperatures below ambient were maintained with a Forma Scientific Inc. “cold finger” in conjunction with the Bronwill circulator. The system was allowed to equilibrate overnight and the temperature inside the dewars varied a maximum of 1”from the start of a run to the end. Plastic syringes containing approximately 0.3-2 mmol of sample (the amount depended upon the value of AR,) were filled in a drybox and sealed with rubber septa. Three syringes were placed into each of two dewar flasks which contained 200 ml of solvent and were allowed to equilibrate. A heating curve was run prior to each injection. and the temperature was measured immediately after injection. After the contents of all six syringes were delivered, six more were placed in the same dewar flasks, and the runs repeated.

Results The partial molal heat of solution values (AITs) at high dilution in water for NaCl and five tetraalkylammonium bromides at a series of temperatures are listed in Tables I-VI. Except for tetrapropylam-

Table ILLLMe4NBr

A P 8 , kcal/mol 5.972 5.921 5.896 5.821 5.789 5.671 5.695 5.560 5.641

Ag,(25-) a Least squares intercept: (AC,,” = -11.3 i 0.1 cal/mol deg. Table i 1 i . O

r , “C

1.723 i 0.043 1.686 i 0.025 1.372 i 0.030 1.331 i 0.017 1.014 i 0.042 0.981 i 0.050 1.007 i 0.067 0.882 i 0.027 0.867 i 0.027 0.667 i 0.030 0.658 f 0.034 0.602 f 0.020 0.561 i 0.036 0.387 i 0.020 0.350 i 0.034

6.1 6.3 14.6 14.8 24.3 24.4 24.6 28.9 29.5 37.5 37.7 39.9 40.6 48.7 48.9

of

the American Chemical Society

92:24

=

5.868 i 0.012 kcal,’mol,

EtlNBr

1.023 0.985 1.341 1.426 1.596 1.625 1.830 1.889 2.123 2.241

I,

A8425‘) a Least squares intercept: ACp,” = 30.0 i 0.9 cal/mol deg. Table i V s U Pr,NBr

=

______

.

0 3 6 7

c

13 13 22 25 76 28

7

s

0

4 0 0

-17 6

46 6 =

- 1.18 I 0.01 kcal/mol:

Table V.R BuaNBr

An,,kcal/mol i 0.12 i 0.19

I 0.184 f 0.151 i 0.06 i 0.120 _i; 0.080 C 0.111 i 0.06

i 0 147

+ 0.090 IC,

0.109

C 0.094 f 0.088 i 0.115

~ l R , ( 2 5 ~=) -2.21 a Least squares intercept: ACp,” = 176 =k 1 cal/mol deg.

1

4

____ I,

35 i 0 13 38 I 0 11 49 i 0 01 20 2 0 06 04 & 0 10 88 i. 0 14 7 7 1 001 16 i 0 10

Least squares intercept: AR,(25 ”) ACP2“= 106 i 1 crll/mol deg.

--

3 9 7 9 8

1.411 1- 0.013 kcal/mol;

ARa,kcalimol -2 -2 -1 -1 -1 -0 0 1

“C

10 IO 23 26 30 32 40 40 49 50

0.073 =t0.085 i 0.018 0.038 It 0.024 i 0.031 i 0.065 =t 0.086 =t0.101 i 0.065 k

-5.64 -5.48 -3.86 -3.68 -2.93 -2.26 -2.18 -1.99 -1.62 0.05 0.10 1.47 1.63 3.62 3.89

(11) C. N. Reilley and D. 7.Sawyer, “Experiments for Instrumental Methods,” McGraw-Hill, New York, N. Y . , 1961. (12) L. Meites and H. C. Thomas, “Advanced Analytical Chemistry,” McGraw-Hill, New York, N. Y . , 1958. Journal

13.6 22.8 23.0 29.9 31 .o 40.9 42.5 17.9 4 . 2

Anq,kcal/mol

Table I. NaCl

A P s , kcalimol

t . ‘C

0.098 0.015 0.026 0.099 j~0.034 i 0.076 i 0.058 .* 0.055 + 0.091 zt i zt d=

t , ”C

___-

5.7 5.8 16.5 16.6 20.6

25.3 25.4 26.5 21.5 37.7 37.8 46.2 47.0 58.2 59.5 3:

R02 kcal/mol;

monium bromide, our results for S C p z uagree within combined experiinental error with those of Sarma, et a/.,lo although their values and errors are generally higher than ours. Our reported error limlts are the standard dekiations for six mensureinents at each

December 2, 1970

7099 I

I

2 000

k\

n

4001

3600-

0200I000l

I0

20

30

40

50

I

60

T e m p e r o ' u r e , 'C

000GL

0"

I

lo"

I

Figure 1. A H , values as a function of temperature for salts in Tables I-VI.

temperature. In Figure 1, these data are summarized and plotted. A linear variation of ARSwith temperature is observed, or, stated differently, the heat capacities of solution (AC,,") are constant within our experimental error over this temperature range. Table V1.a (i-Ar&NF%r

AB,, kcalimol

t , "C

-5.10 i 0 . 1 7 -5.02 f 0 . 2 0 -4.53 f 0.17 -3.77 i. 0 . 2 0 -2.65 i 0.18 -2.66 f 0 . 0 9 -1.55 f 0 . 1 6 -0.229 f 0.275 0.037 f 0.246

17.8 18.3 21.5 24.6 29.3 29.5 33.4 41.1 41.3

a Least squares intercept: AR,(25') ACP," = 219 i 5 cal/mol deg.

=

-3.604 f 0.047 kcal/mol;

The apparent constancy of AC,," is unexpected in view of the work of Ackerman, et who reported definite temperature dependence of the partial molar heat capacities for a series of alkali halides and organic salts. Results by other workers cited in our first paper6 also show considerable temperature dependence of €, obtained from specific heat data. By eq 1, AC,," differs from c,, by the heat capacity of the pure solid (C,'), a nearly constant term for a given salt; hence, the variation of ACP," with temperature should be almost identical with that of c,,. Two possible reasons for our failure t o find a temperature variation for ACpZ0could be (a) that our calorimeter does not measure ARs with precision adequate t o detect nonlinear temperature dependence over this range, or (b) that curvature in Ackerman's results were caused by solute-solute interactions. His method entails the measurement of the specific heats

(cPJ

'

'

20"

-

30' 40' Temoerature. 'C

I

t

50'

1

1 60"

Figure 2. Comparison of A S 8 values for NaCl as a function of temperature: 0,values of Criss and Cobble; 0,from this paper; ----,linear least squares fit of data from this paper.

of relatively concentrated solutions (0.5-2 m ) and an extrapolation to infinite dilution so that temperaturedependent heat capacity of dilution terms could be involved. In view of his difficulty in extrapolating apparent molal heat capacities t o infinite dilution for tetra-n-butylammonium bromide, and the dramatic concentration effects observed by White and Benson l 3 for potassium octanoate, we believe that explanation b (see above) is attractive and reasonable for the larger ions. The results of Sarma, et al.,'" are not relevant t o this point since they were determined at only two temperatures. We have tested the precision of our calorimeter over a 50" temperature range by measuring the heat of solution ol" NaCl as a function of temperature in comparison with the precise study of Criss and Cobble.* The NaCl data in Figure 1 are presented (including the temperature variation) together with Cobble's results in Figure 2 . The curvature of his plot is identical with that of ours. The slightly larger endothermicity of our values relative t o his is due t o a concentration d e p e n d e n ~ e . ' ~Our final concentration was 0.03 M while Cobble's was 0.02 M . The linear temperature dependence of the Me4NBr heats can be shown by a comparison with the NaCl data. Both results were treated with linear and quadratic least squares equations. l j The quadratic fit reduced the standard deviation of the NaCl points (sL = k0.059 us. sa = k0.025 kcal/mol), while the Me4NBr points remained practically unaffected (SL = f0.030 us. sa = *0.029 kcal/mol). The errors for the other salts, especially the three with largest cations, (13) P. White and G. C. Benson, J . Phys. Chem., 64, 599 (1960). (14) V. P. Parker, N.S.R.D.S.-NBSZ, National Bureau

of

Standards, April 1965. (15) W. l . Youden, "Statistical Methods for Chemists," Wiley, New York; N. Y., 1951.

Arnett, Campion

1

Tetraalkylammonium Bromides

7100 +2601

I

-v

-133;

1

/

1

1

,

8

12

I 16

20

24

28

C a r b o ? Number

Figure 3 . Plot of ACp,‘ for straight chain tetraalkylammonium bromides cs. carbon number (NH4Br and (i-Am)rNBr data were not used in drawing the correlation line shown).

are sufficiently great that the plots are experimentally indistinguishable from a straight line. Conspicuously absent from the series reported here are results for tetra-n-amylammonium bromide. This salt dissolves so slowly in water, especially at high temperatures, that reliable enthalpies could not be measured. Discussion Thermodynamic Properties. The process t o which our thermodynamic properties apply is the transfer of 1 niol of solute from its pure liquid, solid, or gaseous state t o high dilution in water a t 25’ (see eq 1). In view of the fact that the salts are all crystalline a t this temperature and that lattice energies are determined by a variety of complex factors,iGthere is an extra degree of uncertainty in interpreting AP,’ (where P stands for any thermodynamic property) for the solution of these solids i n water unless their sublimation properties are known. In contrast, some of the corresponding properties for coninion nonelectrolytes are relatively easy to handle since many of their heats of vaporization are accessible for referring heats of solution back t o a gaseous standard state. Furthermore, in their pure liquid states they are sufficiently similar t o most solvents (with the principle exception of water) that heats of mixirig or heats of solution are usually small. Figure 1 is therefore interesting in view of the sign and magnitude of AITS for the solutes as a function both of molecular structure and of temperature. Obviously, the heats of solution in cool water of the bromides displayed there are related directly to their carbon number. Furthermore, the differentiation between them is dependent on the temperature of the solution, the greatest separation occurring a t lowest temperature. As the temperature is raised, the heats of solution tend t o converge so that in hot water we would expect t o find a much less systematic arrangement of A g s within a smaller range of values. These results suggest some kind of interaction between cold water and the hydrophobic portion of the ions and also suggest that this solvation interaction is disrupted at higher temperatures. The fact that these (16)

s. S. Chang and E. F. Westrum, J . Chem. Phj,s., 36, 2420 (1962).

salts are so widely differentiated a t low temperatures indicates that differences in their lattice energies are either small relative t o their heats of solvation or are proportional t o them. Since the pattern is so similar t o that which has been reported for the alcohols, the prior alternative is more reasonable. We therefore consider that the heat capacities of these salts like those of the corresponding alcohol^^^^ are overwhelmingly dominated by a large exothermic contribution resulting from the interaction of the hydrophobic portion of the carbon chain with water a t low temperature. The nature of the attractions which produce this exothermic development of structure is, of course, related directly t o the whole controversy of the structure of water and how it is modified by solutes. The sign of serves to dichotomize structure-breaking and structure-making solutes in water more sharply than any other therniodynamic property we know. All inorganic salts of which we are aware have negative partial molal heat capacities in water and all nonelectrolytes have positive ones. By this criterion, the tetraalkylamriioniuni salts starting with tetramethylammonium behave like nonelectrolytes and the direct correlation of ACP!O with carbon number is exactly analogous t o that found i n other nonelectrolytes of which the alcohols are the best documented e x a n i p l e ~ . ~ ~ ~ ~ An important practical consideration which follows from the data represented in Figure 1 is the danger of interpreting heats of solution of nonelectrolytes or alkylammonium salts in water at a single temperature. Obviously, the heats of solution and (more important) the differences between them are very sensitive t o temperature as well as t o molecular structure. Concerning the Persistance of Clathrate Structure at High Dilution. Jeffrey and his group have reported a systematic study of clathrate hydrates including some tetraalkylammoniuni salts. l 7 In this series. tetra-nbut ylam m onium ion and t et r ai so amyl ani in on i u ni i on have a particularly strong tendency toward the formation of clathrate hydrates. A number of authors have suggested that some of the properties of nonelectrolytes in aqueous solution might be interpreted in terms of a tendency t o form icelike cages with the geometry of the clathrate even at high dilution i n water. A result which bears on this notion is prokided by comparison of t et r ai so amyl am ni o n iu ni br on1i de, which forms a very stable clathrate at 25”, with the other members of the series, which d o not. One might suppose that the creation of a specially stable icelike cage around this ion at high dilution would lead t o an abnormal ACPlovalue compared t o the other members of the series. This is not observed in Figure 3. This result does not support the “clathrates at high dilution” model but should not be taken as overwhelming evidence against it. The particular stability of the tetraisoamylamnioniuni bromide clathrate in its crystalline form may be due t o some special geometrical contribution t o the lattice energy as repeating units are placed side by side. This type of reinforcement would be lacking at high dilution and thereby remove the singularity in this case. AC,,’ in Ethanol. The large magnitudes of ACi,jo for R,N- salts in water provide strong ebidence for

cp,

(17) G . A . Jeffrey and R. I